When designing observability systems, performance and coverage often take priority over efficiency. Yet every metric scraped, log retained, and trace collected consumes storage, compute, and energy. Sustainability in observability goes beyond reducing costs, it means choosing architectures, data strategies, and tools that minimize environmental impact while maintaining system reliability.
This talk explores how to apply green software principles to observability. We’ll look at practical ways to reduce data bloat, optimize telemetry pipelines, and lower energy use: all through open source technologies. Examples include right-sizing metric retention in Prometheus-compatible systems like VictoriaMetrics, efficient log handling with Loki and Fluent Bit, and adaptive sampling with OpenTelemetry to balance visibility with resource impact.
Real-world scenarios will show how organizations have achieved smaller storage footprints and reduced compute waste by rethinking how and what they monitor. We’ll also discuss how observability practices align with the Green Software Foundation’s principles for carbon awareness, and how teams can measure and report their improvements.
By the end of this session, attendees will understand how to build observability pipelines that are not only scalable and reliable but also environmentally responsible, proving that better observability can go hand in hand with sustainability.
